© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 17501
    Prev
    Next
    INDEX
    Explanations

    words related to political organizations, international relations, and controversial topics

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    480
    +0.38
    1.8%
    481
    +0.15
    0.7%
    447
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.38
    0.08
    146
    +0.15
    0.06
    481
    +0.09
    0.07
    Negative Logits
    rote
    -0.67
    soDeliveryDate
    -0.62
     Mehran
    -0.58
    »Ĵ
    -0.58
    rotein
    -0.56
    eport
    -0.55
    imore
    -0.52
    eatures
    -0.52
    tml
    -0.51
    ãĤ¹ãĥĪ
    -0.50
    POSITIVE LOGITS
     being
    1.23
     becoming
    1.12
     needing
    1.10
     having
    1.05
     wanting
    1.04
     interfering
    1.02
     liking
    0.99
     getting
    0.99
     gaining
    0.97
     failing
    0.97
    Activations Density 0.676%

    No Known Activations